lkml.org 
[lkml]   [2006]   [Nov]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
SubjectRe: + sched-use-tasklet-to-call-balancing.patch added to -mm tree
On Fri, 10 Nov 2006, Siddha, Suresh B wrote:

> Nick was suggesting another alternative, that we should increase the max
> interval and probably make it dependent on number of cpus that
> share the group. Today for NUMA domain, we already set max_interval
> based on online cpus. Perhaps we need to do that for other domains (SMP, MC)
> too and increase the max interval on NUMA for big systems?

I have been trying to run with this patch. We increase the min interval so
that we can at least put one tick between each of the load balance actions
of each processor. The cpu_offset can be modified to:

/* Don't have all balancing operations going off at once: */
static inline unsigned long cpu_offset(int cpu)
{
if (NR_CPUS < HZ)
return jiffies + cpu * HZ / NR_CPUS;
else
return jiffies + cpu;
}

But exclusion in load balancing would take care of the above.




Index: linux-2.6.19-rc4-mm2/include/linux/topology.h
===================================================================
--- linux-2.6.19-rc4-mm2.orig/include/linux/topology.h 2006-11-07 16:43:43.084665449 -0600
+++ linux-2.6.19-rc4-mm2/include/linux/topology.h 2006-11-07 16:55:23.027245226 -0600
@@ -182,7 +182,7 @@
.parent = NULL, \
.child = NULL, \
.groups = NULL, \
- .min_interval = 64, \
+ .min_interval = max(64, jiffies_to_msec(num_online_cpus())),\
.max_interval = 64*num_online_cpus(), \
.busy_factor = 128, \
.imbalance_pct = 133, \
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
\
 
 \ /
  Last update: 2006-11-11 02:17    [W:0.913 / U:0.140 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site